Method and system for predicting task completion of a time period based on task completion rates of prior time periods using machine learning

Abstract

A request is received from a client for determining task completion of a first set of tasks associated with attributes, the first set of tasks scheduled to be performed within a first time period. For each of the attributes, a completion rate of one or more of a second set of tasks is calculated that are associated with the attribute. The second set of tasks has been performed during a second time period in the past. An isotonic regression operation and/or temporal smoothing are performed on the completion rates associated with the attributes of the second set of tasks that have been performed during the second time period to calibrate the completion rates. Possible completion for the attributes of the first set of tasks to be performed in the first time period is calculated based on the calibrated completion rates of the second set of tasks.

What is claimed is: 
     
       1. A computer-implemented method for predicting task completion, the method comprising the operations of:
 receiving a request from a client for predicting task completion rates for a first set tasks, each of the first set of tasks associated with one or more of a plurality of attributes and scheduled to be performed within a first time period, wherein each of the plurality of attributes represents a task processing stage; 
 in response to the request, calculating, for each of the plurality of attributes, a completion rate of one or more of a second set of tasks that are associated with the attribute, wherein the second set of tasks have been performed during a second time period in the past, wherein the completion rate represents a percentage of the one or more tasks associated with the attribute that have been completed during the second timer period, wherein the completion rate for each of the plurality of attributes is smoothed by averaging completion rates of the respective attribute within a plurality of predetermined time windows shifted forward and backward for dates surrounding a given date within the second time period; 
 performing an isotonic regression operation on completion rates associated with the plurality of attributes of the second set of tasks that have been performed during the second time period to calibrate the calculated completion rates, wherein calibrating the calculated completion rates includes adjusting the calculated completion rates against a determined monotonic historical trend of completion rates of the plurality of attributes associated with tasks during a plurality of time periods in the past; 
 determining, for each of the plurality of attributes, one or more tasks from the first set of tasks that are associated with that attribute; 
 predicting a completion rate for the one or more determined tasks associated with the attribute based on a calibrated rate of the completion rates corresponding to that attribute; and 
 iteratively, for a predetermined number of iterations, performing the operations of receiving a request, calculating a completion rate, performing an isotonic regression, determining one or more tasks, and predicting a completion rate; 
 wherein, for each iteration of the predetermined number of iterations, a new set of tasks are used as the first set of tasks; and 
 wherein, for each iteration of the predetermined number of iterations, the first set of tasks and the second set of tasks for a preceding iteration are merged and used as the second set of tasks for that iteration of the predetermined number of iterations. 
 
     
     
       2. The method of  claim 1 , wherein the first time period and the second time period have the same number of days, and wherein the first time period and the second time period end on the same calendar day of a month. 
     
     
       3. The method of  claim 1 , further comprising determining a trend of completion rates for a plurality of tasks associated with the attributes that have been performed in a plurality of time periods in the past, wherein the completion rates of tasks associated with the second time period are calibrated according to the determined trend of completion rates in the past. 
     
     
       4. The method of  claim 3 , wherein performing an isotonic regression operation on the completion rates associated with the attributes comprises modifying at least one of the completion rates according to the trend of completion rates, such that the completion rates of the attributes of the second set of tasks are aligned with the trend of completion rates. 
     
     
       5. The method of  claim 4 , wherein performing an isotonic regression operation on the completion rates associated with the attributes further comprises applying an optimization function to the completion rates. 
     
     
       6. The method of  claim 1 , further comprising calculating completion rates of the attributes for a plurality of past time periods, the second time period being one of the plurality of past time periods, wherein predicting completion rates for the attributes of the first set of tasks is performed based on the completion rates of the plurality of past time periods. 
     
     
       7. The method of  claim 6 , further comprising:
 normalizing opportunity data of the attributes associated with the current time period to generate a current opportunity vector; 
 for each of the past time periods,
 normalizing opportunity data of the attributes corresponding to the past time period to generate a past opportunity vector, 
 determining a weight factor for the past time period based on a comparison between the current opportunity vector and the past opportunity vector; and 
 
 predicting the completion rates for the attributes of the first set of tasks based on completion rates of the past time periods in view of the corresponding weight factors. 
 
     
     
       8. A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations of predicting task completion of a plurality of tasks, the operations comprising the operations of:
 receiving a request from a client for predicting task completion rates for a first set tasks, each of the first set of tasks associated with one or more of a plurality of attributes and scheduled to be performed within a first time period, wherein each of the plurality of attributes represents a task processing stage; 
 in response to the request, calculating, for each of the plurality of attributes, a completion rate of one or more of a second set of tasks that are associated with the attribute, wherein the second set of tasks have been performed during a second time period in the past, wherein the completion rate represents a percentage of the one or more tasks associated with the attribute that have been completed during the second timer period, wherein the completion rate for each of the plurality of attributes is smoothed by averaging completion rates of the respective attribute within a plurality of predetermined time windows shifted forward and backward for dates surrounding a given date within the second time period; 
 performing an isotonic regression operation on the completion rates associated with the plurality of attributes of the second set of tasks that have been performed during the second time period to calibrate the calculated completion rates, wherein calibrating the calculated completion rates includes adjusting the calculated completion rates against a determined monotonic historical trend of completion rates of the plurality of attributes associated with tasks during a plurality of time periods in the past; 
 determining, for each of the plurality of attributes, one or more tasks from the first set of tasks that are associated with that attribute; 
 predicting a completion rate for the one or more determined tasks associated with that attribute based on a calibrated rate of the completion rates corresponding to that attribute; and 
 iteratively, for a predetermined number of iterations, performing the operations of receiving a request, calculating a completion rate, performing an isotonic regression, determining one or more tasks, and predicting a completion rate; 
 wherein, for each iteration of the predetermined number of iterations, a new set of tasks are used as the first set of tasks; and 
 wherein, for each iteration of the predetermined number of iterations, the first set of tasks and the second set of tasks for a preceding iteration are merged and used as the second set of tasks for that iteration of the predetermined number of iterations. 
 
     
     
       9. The machine-readable medium of  claim 8 , wherein the first time period and the second time period have the same number of days, and wherein the first time period and the second time period end on the same calendar day of a month. 
     
     
       10. The machine-readable medium of  claim 8 , wherein the operations further comprise determining a trend of completion rates for a plurality of tasks associated with the attributes that have been performed in a plurality of time periods in the past, wherein the completion rates of tasks associated with the second time period are calibrated according to the determined trend of completion rates in the past. 
     
     
       11. The machine-readable medium of  claim 10 , wherein performing an isotonic regression operation on the completion rates associated with the attributes comprises modifying at least one of the completion rates according to the trend of completion rates, such that the completion rates of the attributes of the second set of tasks are aligned with the trend of completion rates. 
     
     
       12. The machine-readable medium of  claim 11 , wherein performing an isotonic regression operation on the completion rates associated with the attributes further comprises applying an optimization function to the completion rates. 
     
     
       13. The machine-readable medium of  claim 8 , wherein the operations further comprise calculating completion rates of the attributes for a plurality of past time periods, the second time period being one of the plurality of past time periods, wherein predicting completion rates for the attributes of the first set of tasks is performed based on the completion rates of the plurality of past time periods. 
     
     
       14. The machine-readable medium of  claim 13 , wherein the operations further comprise:
 normalizing opportunity data of the attributes associated with the current time period to generate a current opportunity vector; 
 for each of the past time periods,
 normalizing opportunity data of the attributes corresponding to the past time period to generate a past opportunity vector, 
 determining a weight factor for the past time period based on a comparison between the current opportunity vector and the past opportunity vector; and 
 
 predicting the completion rates for the attributes of the first set of tasks based on completion rates of the past time periods in view of the corresponding weight factors. 
 
     
     
       15. A data processing system, comprising:
 a processor; and 
 a memory coupled to the processor to store instructions, which when executed by the processor, cause the processor to perform operations of predicting task completion of a plurality of tasks, the operations including
 receiving a request from a client for predicting task completion rates for a first set tasks, each of the first set of tasks associated with one or more of a plurality of attributes and scheduled to be performed within a first time period, wherein each of the plurality of attributes represents a task processing stage; 
 in response to the request, calculating, for each of the plurality of attributes, a completion rate of one or more of a second set of tasks that are associated with the attribute, wherein the second set of tasks have been performed during a second time period in the past, wherein the completion rate represents a percentage of the one or more tasks associated with the attribute that have been completed during the second timer period, wherein the completion rate for each of the plurality of attributes is smoothed by averaging completion rates of the respective attribute within a plurality of predetermined time windows shifted forward and backward for dates surrounding a given date within the second time period; 
 performing an isotonic regression operation on the completion rates associated with the plurality of attributes of the second set of tasks that have been performed during the second time period to calibrate the calculated completion rates, wherein calibrating the calculated completion rates includes adjusting the calculated completion rates against a determined monotonic historical trend of completion rates of the plurality of attributes associated with tasks during a plurality of time periods in the past; 
 determining, for each of the plurality of attributes, one or more tasks from the first set of tasks that are associated with that attribute; 
 predicting a completion rate for the one or more determined tasks associated with that attribute based on a calibrated rate of the completion rates corresponding to that attribute; and 
 iteratively, for a predetermined number of iterations, performing the operations of receiving a request, calculating a completion rate, performing an isotonic regression, determining one or more tasks, and predicting a completion rate; 
 wherein, for each iteration of the predetermined number of iterations, a new set of tasks are used as the first set of tasks; and 
 wherein, for each iteration of the predetermined number of iterations, the first set of tasks and the second set of tasks for a preceding iteration are merged and used as the second set of tasks for that iteration of the predetermined number of iterations. 
 
 
     
     
       16. The system of  claim 15 , wherein the first time period and the second time period have the same number of days, and wherein the first time period and the second time period end on the same calendar day of a month. 
     
     
       17. The system of  claim 15 , wherein the operations further comprise determining a trend of completion rates for a plurality of tasks associated with the attributes that have been performed in a plurality of time periods in the past, wherein the completion rates of tasks associated with the second time period are calibrated according to the determined trend of completion rates in the past. 
     
     
       18. The system of  claim 17 , wherein performing an isotonic regression operation on the completion rates associated with the attributes comprises modifying at least one of the completion rates according to the trend of completion rates, such that the completion rates of the attributes of the second set of tasks are aligned with the trend of completion rates. 
     
     
       19. The system of  claim 18 , wherein performing an isotonic regression operation on the completion rates associated with the attributes further comprises applying an optimization function to the completion rates.